Do not apply additional encoding when reading process output.
Fixes errors when encountering UTF-8:
Traceback (most recent call last):
File "./generate_expected_output.py", line 117, in <module>
generateTestData(path)
File "./generate_expected_output.py", line 106, in generateTestData
out.write(data.decode('utf-8'))
UnicodeEncodeError: 'ascii' codec can't encode character u'\xdc' in position 5485: ordinal not in range(128)
Change-Id: Ib827787a59a18b4d3d0601645856517f43c01fc3
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
Evaluate command line arguments and use directories only when empty.
Change-Id: I818ec13c686018a3f607e91174e57d8f8bbf15f7
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@digia.com>
The actual code in testlib now indents the duration tag when it is for a
function.
Change-Id: Iee62db9c81f11dc54e57f166bf9fb2b7012b7e03
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
This makes testlib a lot more hackable since when the output is changed
in a defined way we can regenerate the expected output (eg adding the
test duration which will be a follow-up patch).
Note that the script does not work properly on Windows and not all
benchmarks work, but at least it reduces the work of adapting files to a few lines.
Change-Id: I910387cd92ff82aa629747a3a3033dae17fbd711
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
